Visible Light Positioning (VLP)

My Ph.D. research work is related to indoor localization using visible light emitters. The 3-D model of an empty room is developed with a certain number of LED light sources and receivers. The light sources are considered lambertian with lambertian index 'm' and receivers are assumed to have a certain field-of-view (FOV). The work is still in progress. Updates will come soon.

Indoor Localization using Wi-Fi Fingerprinting

My MS Thesis research work is related to indoor localization problem. My work, in particular, makes use of Received Signal Strength (RSS) from WiFi Access Points (APs). The brief overview of my research work is following.

The main limitation of deploying/updating Received Signal Strength (RSS) based indoor localization system is the construction of fingerprinted radio map, which is quite a hectic and time-consuming process especially when the indoor area is enormous and/or dynamic. The performance degrades when the fingerprinting load is reduced below a certain level. We proposed an indoor localization scheme that requires very small fraction of fingerprinting load (1% of total grid points, i.e. 2 out of 219 points in our case), some crowd sourced readings and plan coordinates of the indoor environment. The 1% fingerprinting load is used only to perturb the local geometries in the plan coordinates. The concept of geometry perturbation is shown in the figures at right (Top: without geometry perturbation and Bottom: with geometry perturbation). Our proposed algorithm is shown to achieve less than 5m mean localization error with 1% fingerprinting load and a limited number of crowd sourced readings, when other learning based localization schemes pass the 10m mean error with the same information. The performance is further improved by clustering the crowd sourced information, where the few collected fingerprints in the offline phase act as cluster heads.The few location estimations together with few fingerprints help to estimate the complete radio map of the indoor environment. The estimation of radio map does not demand extra workload rather it employs the already available information from the proposed indoor localization framework. The testing results for radio map estimation show almost 50% performance improvement by using the aforementioned information as compared to using only fingerprints.

2-D Direction-of-Arrival (DOA) Estimation

We proposed 2D DOA estimation algorithm using L-shaped array. The proposed technique makes use of cross-correlation information of the signals from non-coherent sources at the antenna arrays. The proposed method is also effective in estimating the elevation and azimuth angles with no failure in estimation particularly in the mobile application range, as well as much lower RMSE. The performance is improved and complexity is greatly reduced as compared to the existing schemes.